Maps help navigators determine their position in the world. We define navigational checking as an aviator's task of continually comparing a map with the forward field of view (FFOV). We report two experiments using static three-dimensional (3D) images as the map and FFOV. Participants were asked to judge as quickly and accurately as possible whether the map and FFOV matched or mismatched. Elevation angle, azimuth angle, type of mismatch trial, and image complexity were manipulated; both response time and accuracy were recorded. Results were generally as predicted. Of greatest interest, we found a fairly linear performance decrement associated with the difference in foreshortening between the two views (a trigonometric conversion of elevation angle disparity), a somewhat nonlinear performance decrement associated with azimuth angle disparity, and no interaction between the two. Finally, we address how elevation and azimuth might influence 3D map display design.

In three experiments, subjects performed a navigational checking task in which the view presented on an electronic map was compared with the view of a simulated world to determine the congruence. Map viewpoints were varied in elevation and azimuth angIe disparity relative to the world view to simulate the effects of map rotation and 3D map angle on navigational checking. On most trials the area depicted was the same, whereas on a small percentage, features of the map were altered, requiring the subject to judge "mismatch." In Experiment 1, using simple images, response time increased non-linearly with elevation angle distortion. In Experiment 3, using more complex images, similar effects were found and were augmented by effects of azimuth angle distortion. In Experiment 3, using dynamic realistic real world scenes, elevation angle and map scale effects were examined and revealed a complex pattern. The results are interpreted in terms of guidelines for 3D electronic map construction.A critical facet of a pilot's flying task in visual meteorological conditions is navigational checking. This is the process of maintaining a fairly continuous comparison between geographical features in the forward field of view (FFOV), and some representation of those features on a map, in order to determine that the two are congruent (i.e., what I see, is what I should see; Are& 1991). The occurrence of breakdowns in this process is evident, for example, in the case of wrong airport landings (Antunano et al., 1989), many of which occur in conditions of good visibility. The physical nature of the map representation becomes particularly important on less standardized missions, such as military combat sorties (Kibbe, and Stiff, 1993), search and rescue operations, or EMS evacuations. In such cases time is critical, the consequences of navigational errors are severe, and automated navigational aides are not always available.The navigational checking task of comparing two images to determine their congruence bears a parallel to various "same-different" judgment tasks examined in cognitive psychology (Posner, 1978;Cooper and Podgorny, 1976); however, in the former case the images are far more complex, and the concept of "sameness" is one that is explicitly defined by the rule: "is the region depicted on the map, the same as that depicted in the FFOV," independent of other physical differences between the two images?As in the studies carried out in simpler laboratory paradigms, we assume that "same'' judgments will be made most rapidly when there is actual physical congruence between the two images (Cooper and Shepard, 1978). That is, the map image is a virtual photographic rendering of the momentary and evolving FFOV. However, for an in-flight electronic map, this circumstance would require a high fidelity image generator, constantly updating the representation of the forward three-dimensional view, as the aircraft traverses the terrain or approaches a point on the ground. The expense of such a system is evident, and if data must be update...